Several randomized controlled trials have demonstrated that exercise can benefit cancer patients in terms of their physical functioning and symptoms. However, few studies have included patients with advanced disease, a group that is living longer with their disease as more treatments are developed. For example, with the recent development of multiple novel agents in the treatment of metastatic colorectal cancer, the overall survival of these patients has nearly tripled. Exercise may help these patients maintain function and manage symptoms, and facilitate their continuing therapy for as long as it is tolerated and effective in controlling disease. The goal of the proposed study is to assess the feasibility and develop preliminary estimates of effect size of a home- based exercise intervention on physical functioning of patients with metastatic and/or recurrent colon cancer in the setting of both a comprehensive cancer center and the community clinical oncology practice network (CCOP). This R21 is proposed to collect data to help plan for a large trial of the intervention. The specific aims of this R21 address three areas related to planning a larger trial: (1) Effect sizes: We aim to develop estimates of the effect size of the exercise intervention on the primary outcome of physical functioning (self-reported and performance tests) and the secondary outcomes of symptoms (fatigue, sleep, pain, nausea, psychological distress) of patients with advanced colon cancer, compared to an attention control intervention (relaxation training);(2) Feasibility: The proposed research will evaluate study feasibility, operationalized as patient adherence, program implementation, contamination (e.g., intervention given or extent of exercise in control group), patient satisfaction, and implementation complexity as rated by clinic staff. (3) Theory: To explore Social Cognitive Theory-based model of predictors of adherence to exercise. We will evaluate psychometric characteristics of the measures of these variables and estimate the effects of the exercise intervention on the hypothesized mediators, as well as explore associations between the model variables and exercise adherence. This study will recruit patients with unresectable stage IV or recurrent colon cancer (ECOG performance status 0-1) who have started a new chemotherapy regimen in the past 28 days. Participants will be randomized to one of two arms: (1) Home-based exercise including pedometers for monitoring walking and resistance bands for functional strength training;(2) Relaxation training using diaphragmatic breathing and guided visualization. The relaxation training arm is included as an attention control, since many of the outcomes for the proposed study, as well as any potential future trials, will rely on self-report. Both study arms will receive print materials and telephone counseling from M. D. Anderson study staff to encourage adherence to the behavior, as well as a letter and brief advice from the patient's oncologist about the importance of exercise or relaxation training. Participants will complete questionnaires when they enter the study and after 16 weeks of intervention. They also will undergo brief physical performance tests in the clinic at baseline and at 16 weeks. PUBLIC HEALTH RELEVANCE: The goal of this study is to gather preliminary information on the effectiveness and feasibility of an exercise intervention for patients with advanced colon cancer. We will assess whether an exercise intervention in which patients receive telephone counseling, advice from their oncologist, written materials, DVD instructions, and a pedometer and resistance bands improves the physical functioning and symptoms of patients, compared to a relaxation training intervention. If this intervention is effective, it could be provided to patients through their oncologist's office, and may help them maintain activities of daily living and independence for longer periods of time during treatment.